Piperacillin-tazobactam (P/T) is a β-lactam–β-lactamase inhibitor combination frequently used in the hospital setting. Etest is a gradient diffusion method that represents an alternative to broth microdilution (BMD) for performing antimicrobial susceptibility testing. We conducted a multicenter evaluation of the performance of the new P/T Etest compared to that of BMD following U.S.
KEYWORDS: piperacillin-tazobactam, Etest, antimicrobial susceptibility testing, gradient diffusion, Enterobacterales, Pseudomonas aeruginosa, Acinetobacter baumannii complex
ABSTRACT
Piperacillin-tazobactam (P/T) is a β-lactam–β-lactamase inhibitor combination frequently used in the hospital setting. Etest is a gradient diffusion method that represents an alternative to broth microdilution (BMD) for performing antimicrobial susceptibility testing. We conducted a multicenter evaluation of the performance of the new P/T Etest compared to that of BMD following U.S. Food and Drug Administration (FDA) and International Standards Organization (ISO) standard ISO 20776-2 criteria using Clinical and Laboratory Standards Institute (CLSI)-FDA and European Committee on Antimicrobial Susceptibility Testing (EUCAST) interpretive breakpoints, respectively. A total of 977 isolates (775 Enterobacterales isolates, 119 Pseudomonas aeruginosa isolates, and 83 Acinetobacter baumannii complex isolates) were tested. Overall essential agreement (EA) was 96.4% and 96.6% for Enterobacterales when FDA and ISO 20776-2 criteria, respectively, were followed. EA was 98.3% for P. aeruginosa and 91.6% for the A. baumannii complex when both the FDA and ISO criteria were followed. Applying CLSI-FDA breakpoints, categorical agreement (CA) reached 93.0%, 93.3%, and 89.2% for the Enterobacterales, P. aeruginosa, and the A. baumannii complex, respectively. Two very major errors (VMEs; 1.1%) were found among the Enterobacterales (for 2 Klebsiella pneumoniae isolates). No additional major errors (MEs) or VMEs were found. Applying EUCAST breakpoints, CA was 94.8% and 95.8% for Enterobacterales and P. aeruginosa, respectively (no breakpoints are currently available for the A. baumannii complex). No VMEs were observed among the Enterobacterales, but 2 (0.4%) MEs were found. Among the P. aeruginosa isolates, 2 (6.9%) VMEs and 3 (3.3%) MEs were observed. These errors resulted when P/T Etest MICs were 1 doubling dilution apart from the BMD MICs. In conclusion, the new P/T Etest represents an accurate tool for performing antimicrobial susceptibility testing of Enterobacterales, P. aeruginosa, and A. baumannii complex isolates with limited category errors.
INTRODUCTION
Piperacillin-tazobactam (P/T) is a β-lactam–β-lactamase inhibitor combination frequently used in the routine clinical setting. Piperacillin is a semisynthetic ureidopenicillin that has activity against both Gram-negative and Gram-positive bacteria but that has a reduced efficacy in the presence of different β-lactamases. Tazobactam, a penicillanic acid β-lactamase inhibitor, protects the activity of piperacillin against Gram-negative isolates producing narrow-spectrum or some extended-spectrum β-lactamases (ESBL) (1–3). Due to the worldwide dispersion of ESBL producers, piperacillin-tazobactam has been viewed as a carbapenem-sparing option (4–6), although controversial reports have also been reported (7, 8).
Although broth microdilution (BMD) is the gold standard method to perform antimicrobial susceptibility testing (9), it is rarely performed in most clinical settings as it is labor-intensive and time-consuming and requires specialized expertise. In addition, several commercial automated susceptibility testing systems as well as disk diffusion are used in routine clinical microbiology laboratories. Easy-to-perform, ancillary tests may also be used to obtain MIC values. Among those, gradient diffusion methods were developed for different antimicrobials, including piperacillin-tazobactam (10). However, in 2015 the European Committee on Antimicrobial Susceptibility Testing (EUCAST) published a warning about problems with piperacillin-tazobactam gradient tests, including Etest PTc (bioMérieux, France). The results of that test were reported to be unreliable due to high variations in MIC values with high category error rates (http://www.eucast.org/ast_of_bacteria/warnings/). Concomitantly, the Etest PTc strip was withdrawn from the market.
We report here the results of an international multicenter evaluation of the recently redesigned piperacillin-tazobactam gradient diffusion strip (P/T Etest; bioMérieux, Marcy l’Etoile, France), approved by the European Union (CE marked in December 2018) and the U.S. Food and Drug Administration (FDA) (January 2019). The aim was to test the equivalence of its performance to that of the reference BMD method to determine the antimicrobial susceptibilities of Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii complex isolates.
MATERIALS AND METHODS
Ethics statement.
Prior to study initiation, each study site acquired approval or a waiver from the respective institutional review board.
Study setting and design.
The performance of the P/T Etest was compared to that of the broth microdilution (BMD) reference method following Clinical and Laboratory Standards Institute (CLSI) M07-A10 (11) and International Standards Organization (ISO) ISO 20776-1 (9) standards. The study design included four performance components: (i) a challenge study, (ii) a clinical study, (iii) a quality control (QC) study, and (iv) a reproducibility study. Inoculum purity was verified for all isolates tested, regardless of the study component. Inoculum density was checked for all quality control replicates, all reproducibility tests, and 10% of the fresh clinical isolates following FDA guidance (12).
The study was conducted between June and October 2018 at 4 different sites, including 3 sites in the United States and 1 site in Europe. The roles of the different sites were as follows: the University Hospital Ramón y Cajal, Madrid, Spain (HRC), performed a clinical study and a quality control study; the University of Rochester Medical Center, Rochester, NY, USA (Rochester), performed a clinical study, quality control study, and reproducibility study; the Washington University in Saint Louis School of Medicine, Saint Louis, MO, USA (WashU), performed a clinical study, quality control study, and reproducibility study; and the Clinical Affairs Laboratory, Saint Louis bioMérieux, Saint Louis, MO, USA (STL), performed a challenge study, clinical study, quality control study, and reproducibility study.
Challenge and clinical studies.
The challenge set was composed of 89 stock isolates. Between June and October 2018, the clinical study recruited 888 clinical isolates, including 448 (50.5%) fresh isolates (isolates tested by up to 7 days from the time isolation and never frozen), 278 (31.3%) recent isolates (isolates tested between >7 days and 1 year from the time of isolation and stored frozen), and 162 (18.2%) stock isolates (isolates tested from >1 to 3 years from the time of isolation). Isolates were identified using matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry according to each site’s internal procedures. No duplicate isolates from the same patient were included. A single 0.5 McFarland standard suspension of each isolate was used to perform the P/T Etest and BMD.
For the P/T Etest, Mueller-Hinton agar plates (BD, Franklin Lakes, NJ, USA) were inoculated by use of a sterile cotton swab moistened with the 0.5 McFarland standard suspension, and the P/T Etest strip was placed on the plate with an applicator or forceps. The P/T Etest result was read according to the manufacturer’s instructions, also included in the package insert, after 16 to 20 h of incubation at 35 ± 2°C (20 to 24 h for A. baumannii). The MIC was read at the location of the complete inhibition of growth. An MIC falling between 2 standard doubling dilutions was rounded up to the next highest value before categorization (Fig. 1). Haze and macrocolonies or microcolonies within 3 mm from the strip were read as growth (Fig. 1).
FIG 1.
Characteristic ellipses observed with QC strains and examples of the results obtained with clinical isolates and the related MICs. (A) QC strain E. coli ATCC 25922 (MIC = 2 μg/ml); (B) QC strain P. aeruginosa ATCC 27853 (MIC = 3 μg/ml); (C) K. pneumoniae clinical isolate (MIC = 3 μg/ml) with a pear-shaped ellipse; (D) K. pneumoniae clinical isolate (MIC = 128 μg/ml) displaying macrocolonies.
BMD was performed in frozen 96-well plates prepared in bioMérieux facilities (La Balme les Grottes, France) in compliance with the directions for the CLSI M07-A10 (11) and ISO 20776-1 (9) standards. Each batch produced was controlled by inoculating several panels withdrawn at the beginning, at the middle, and at the end of production with the QC strains recommended by the CLSI M100 standard (13). The panels were then frozen at −80°C and shipped in dry ice with constant monitoring of the temperature during transportation at all clinical trial sites. The P/T MIC range tested was 0.016 to 256 mg/liter.
Performance was expressed by the use of the data from the clinical and challenge studies combined. MIC values were interpreted using both CLSI and EUCAST breakpoints (13, 14, 15) (Table 1). Performance was expressed using the following indices: essential agreement (EA), categorical agreement (CA), minor error (mE) rate, major error (ME) rate, and very major error (VME) rate. These indices were then compared to the performance criteria defined by the FDA (12) and ISO 20776-2 (16) standards. The performances obtained using EUCAST breakpoints are presented after resolution of the VMEs according to section 5.7.2 of the ISO 20776-2 standard (16). Briefly, for EUCAST-ISO analysis, resolution of MEs and VMEs was attempted by a triplicate test of the reference method only. The consensual BMD mode/median result replaced the initial BMD result. If no consensus result emerged from the triplicate BMD, the isolate was excluded from the analysis. For CLSI-FDA analysis, no resolution of discrepant results was performed.
TABLE 1.
Piperacillin-tazobactam breakpointsa
| Organism | CLSI breakpoint (mg/liter) for: |
EUCAST breakpoint (mg/liter) for: |
||||
|---|---|---|---|---|---|---|
| S | I | R | S | I | R | |
| Enterobacterales | ≤16/4 | 32/4–64/4 | ≥128/4 | ≤8/4 | 16/4 | >16/4 |
| P. aeruginosa | ≤16/4 | 32/4–64/4 | ≥128/4 | ≤16/4 | NA | >16/4 |
| A. baumannii complex | ≤16/4 | 32/4–64/4 | ≥128/4 | IE | IE | IE |
Abbreviations: S, susceptible; I, intermediate; R, resistant; NA, not applicable; IE, insufficient evidence that the organism group is a good target with the agent. Values given before and after a slash are for piperacillin and tazobactam, respectively.
Reproducibility study.
Twelve isolates were included in the reproducibility study, and testing was performed at three sites. Each isolate was tested in triplicate each day on three different days. Three separate 0.5 McFarland standard suspensions of each isolate were prepared in normal sterile saline on each day of reproducibility testing. Each suspension was used for the P/T Etest procedure. The results of the testing were used to compute a modal value for each reproducibility study isolate (if there was no mode, the median was used). Results that were within 1 doubling dilution of the mode were within agreement and were used to calculate the percentage of isolates for which the results were within agreement. Best-case calculation for reproducibility assumed that the off-scale values were within 1 doubling dilution of the mode. Worst-case calculation assumed that off-scale values were not within 1 doubling dilution of the mode. Reproducibility was calculated as the ratio between the number of results with essential agreement over the total number of results. Both within- and between-site reproducibilities are expressed as percentages.
Quality control.
Tests with the quality control isolates for the P/T Etest and BMD methods were conducted at each study site on each day of testing using the following organisms: Escherichia coli ATCC 25922, E. coli ATCC 35218, Klebsiella pneumoniae ATCC 700603, P. aeruginosa ATCC 27853, Enterococcus faecalis ATCC 29212 (BMD only), and Staphylococcus aureus ATCC 29213 (BMD only). Each site tested each quality control strain a minimum of 20 times.
Data analysis.
Results for all isolates tested for P/T susceptibility using the Etest and BMD methods were included in the analysis. Essential agreement (EA) was defined as agreement between the two methods ±1 doubling dilution. Categorical agreement (CA) between the Etest and BMD methods, as well as VME, ME, and minor error evaluations, was determined using CLSI or EUCAST breakpoints.
RESULTS
Challenge and clinical studies.
A total of 89 isolates, including 67 Enterobacterales, 13 P. aeruginosa, and 9 A. baumannii isolates, were used in the challenge study (the composition of the challenge set appears in Table S1 in the supplemental material). In the clinical study, 888 clinical isolates were tested among the 4 sites (242 at HRC, 263 at Rochester, 307 at WashU, and 76 at STL). This population included 708 Enterobacterales, 106 P. aeruginosa, and 74 A. baumannii complex isolates. The overall distribution of these clinical isolates by source of infection was as follows: urine, 45.6%; respiratory tract, 13.3%; blood, 9.5%; wounds, 8.2%; abscesses, 3.9%; body fluid, 2.6%; skin, 2.3%; swabs, 2.3%; feces, 2.0%; tissue, 1.8%; bone, 1.3%; foot, 0.8%; gastric, 0.8%; and others (sample types from which ≤5 isolates each were obtained), 5.6%.
Evaluation of the performance of the P/T Etest included data for 977 isolates from the challenge and clinical studies combined. The distribution of species, susceptibility to P/T, and performance determined using EUCAST or CLSI breakpoints are shown in Tables 2 and 3. The rates of resistance of the isolates to P/T by Etest for Enterobacterales were 22.6% and 29.2%, according to CLSI and EUCAST breakpoints, respectively. Among the P. aeruginosa isolates, resistant isolates accounted for 14.3% and 25.2% of the isolates according to the CLSI and EUCAST breakpoints, respectively. Resistant A. baumannii complex isolates represented 38.5% of the A. baumannii complex isolates tested using CLSI criteria.
TABLE 2.
Overall performance of the new P/T Etest categorized according to CLSI-FDA breakpointsa
| Breakpoint and organism | No. (%) of isolates with the CLSI-FDA breakpoint for: |
No. (%) of isolates with the following CLSI-FDA performance: |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Total | S | I | R | EA | CA | ME | VME | mE | |
| Enterobacterales | 775 (79.3) | 550 (70.8) | 51 (6.6) | 174 (22.6) | 747 (96.4) | 721 (93.0) | 0 | 2 (1.1) | 52 (6.7) |
| Citrobacter freundii | 41 (4.2) | 24 (58.5) | 5 (12.2) | 12 (29.7) | 41 (100) | 37 (90.2) | 0 | 0 | 4 (9.8) |
| Citrobacter koseri | 46 (4.7) | 35 (76.1) | 2 (4.3) | 9 (19.6) | 46 (100) | 45 (97.8) | 0 | 0 | 1 (2.2) |
| Enterobacter cloacae | 36 (3.7) | 22 (61.1) | 3 (8.3) | 11 (30.6) | 36 (100) | 34 (94.4) | 0 | 0 | 2 (5.6) |
| Escherichia coli | 168 (17.2) | 137 (81.6) | 15 (8.9) | 16 (9.5) | 161 (95.8) | 153 (91.1) | 0 | 0 | 15 (8.9) |
| Klebsiella aerogenes | 39 (4.0) | 29 (74.4) | 6 (15.4) | 4 (10.3) | 39 (100) | 35 (89.7) | 0 | 0 | 4 (10.3) |
| Klebsiella pneumoniae | 190 (19.5) | 79 (41.5) | 12 (6.3) | 99 (52.6) | 179 (94.2) | 169 (88.9) | 0 | 2 (2.0) | 19 (10.0) |
| Morganella morganii | 41 (4.2) | 37 (90.2) | 0 | 4 (9.8) | 36 (87.8) | 40 (97.6) | 0 | 0 | 1 (2.4) |
| Proteus mirabilis | 41 (4.2) | 38 (92.7) | 2 (4.9) | 1 (2.4) | 40 (97.6) | 40 (97.6) | 0 | 0 | 1 (2.4) |
| Proteus vulgaris | 31 (3.2) | 100 (100.0) | 0 | 0 | 29 (93.5) | 31 (100) | 0 | 0 | |
| Providencia rettgeri | 28 (2.9) | 22 (78.6) | 0 | 6 (21.4) | 27 (96.4) | 28 (100) | 0 | 0 | 0 |
| Providencia stuartii | 36 (3.7) | 31 (86.1) | 1 (2.8) | 4 (11.1) | 36 (100) | 35 (97.2) | 0 | 0 | 1 (2.8) |
| Salmonella enterica | 31 (3.2) | 31 (100) | 0 | 0 | 31 (100) | 31 (100) | 0 | 0 | |
| Serratia marcescens | 47 (4.8) | 34 (72.3) | 5 (10.6) | 8 (17.0) | 46 (97.9) | 43 (91.5) | 0 | 0 | 4 (8.5) |
| Pseudomonas aeruginosa | 119 (12.2) | 89 (74.8) | 13 (10.9) | 17 (14.3) | 117 (98.3) | 111 (93.3) | 0 | 0 | 8 (6.7) |
| Acinetobacter baumannii complex | 83 (8.5) | 43 (51.8) | 8 (9.6) | 32 (38.5) | 76 (91.6) | 74 (89.2) | 0 | 0 | 9 (10.8) |
S, susceptible; I, intermediate; R, resistant.
TABLE 3.
Overall performance of the new P/T Etest categorized according to EUCAST breakpointsb
| Organism | No. (%) of isolates with the EUCAST breakpoint for: |
No. (%) of isolates with the following EUCAST-ISO performance: |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Total | S | I | R | EA | CA | ME | VME | mE | |
| Enterobacterales | 773a (79.3) | 525 (67.7) | 24 (3.1) | 224 (29.2) | 747 (96.6) | 733 (94.8) | 2 (0.4) | 0 | 38 (4.9) |
| Citrobacter freundii | 41 (4.2) | 23 (56.1) | 1 (2.4) | 17 (41.5) | 41 (100) | 40 (97.6) | 0 | 0 | 1 (2.4) |
| Citrobacter koseri | 46 (4.7) | 31 (67.4) | 4 (8.7) | 11 (23.9) | 46 (100) | 44 (95.7) | 0 | 0 | 2 (4.3) |
| Enterobacter cloacae | 36 (3.7) | 22 (61.1) | 0 | 14 (38.9) | 36 (100) | 34 (94.4) | 0 | 0 | 2 (5.6) |
| Escherichia coli | 167 (17.2) | 131 (78.6) | 5 (2.9) | 31 (18.4) | 161 (96.4) | 158 (94.6) | 1 (0.8) | 0 | 8 (4.8) |
| Klebsiella aerogenes | 39 (4.0) | 26 (66.7) | 3 (7.7) | 10 (25.6) | 39 (100) | 34 (87.2) | 0 | 0 | 5 (12.8) |
| Klebsiella pneumoniae | 189 (19.5) | 72 (37.4) | 7 (3.7) | 110 (58.9) | 179 (94.7) | 175 (92.6) | 0 | 0 | 14 (7.4) |
| Morganella morganii | 41 (4.2) | 37 (90.2) | 0 | 4 (9.8) | 36 (87.8) | 39 (95.1) | 1 | 0 | 1 (2.4) |
| Proteus mirabilis | 41 (4.2) | 37 (90.2) | 1 (2.4) | 3 (7.3) | 40 (97.6) | 40 (97.6) | 0 | 0 | 1 (2.4) |
| Proteus vulgaris | 31 (3.2) | 31 (100) | 0 | 0 | 29 (93.5) | 31 (100) | 0 | 0 | |
| Providencia rettgeri | 28 (2.9) | 22 (78.6) | 0 | 6 (21.4) | 27 (96.4) | 28 (100) | 0 | 0 | 0 |
| Providencia stuartii | 36 (3.7) | 30 (83.3) | 1 (2.8) | 5 (13.9) | 36 (100) | 33 (91.7) | 0 | 0 | 3 (8.3) |
| Salmonella enterica | 31 (3.2) | 31 (100) | 0 | 0 | 31 (100) | 31 (100) | 0 | 0 | |
| Serratia marcescens | 47 (4.8) | 32 (68.1) | 2 (4.3) | 13 (27.7) | 46 (97.9) | 46 (97.9) | 0 | 0 | 1 (2.1) |
| Pseudomonas aeruginosa | 119 (12.2) | 90 (74.8) | 29 (25.2) | 119 (98.3) | 114 (95.8) | 3 (3.3) | 2 (6.9) | 0 | |
| Acinetobacter baumannii complex | 83 (8.5) | 76 (91.6) | NA | NA | NA | NA | |||
One E. coli isolate and one K. pneumoniae isolate were excluded from the EUCAST-ISO evaluation because of high variability in BMD results per ISO 20776-2 standard directions.
S, susceptible; I, intermediate; R, resistant; NA, not applicable.
Performance compared to FDA requirements.
Based on FDA requirements (≥90.0% EA and CA, ≤2.0% VMEs, ≤3.0% MEs, ≥95.0% reproducibility, results for ≥95.0% of the QC isolates within the expected range) (12), we evaluated the performance of the P/T Etest using CLSI breakpoints (Table 2). For the Enterobacterales, overall EA and CA were 96.4% (747/775) and 93.0% (721/775), respectively (Fig. 2A). Two VMEs were observed among the K. pneumoniae clinical isolates; one isolate had an MIC of 16 μg/ml by the P/T Etest and an MIC of 128 μg/ml by BMD, and one isolate yielded an MIC of 8 mg/liter by the P/T Etest and an MIC of 256 mg/liter by BMD. These two isolates were retested for information. For the first isolate, the retest in triplicate resulted in 3 P/T Etest MIC results at 16 μg/ml and 3 BMD MIC results at 32 μg/ml, resolving the error. Regarding the second isolate, a total of 4 triplicate retests were performed because multiple skipped wells invalidated the run several times. Among these 12 replicates, Etest provided an MIC at 16 μg/ml 11 times and an MIC at 32 μg/ml 1 time. When interpretable (i.e., 1 skipped well maximum), the MIC ranged from 8 to 256 μg/ml. For this second isolate, a reference MIC value could not be confidently established due to extensive variability of the results of the reference method. However, as FDA does not allow categorical error resolution, the initial results were kept for the CLSI-FDA analysis. Nevertheless, the P/T Etest met the criteria for VMEs for both the Enterobacterales (1.1%, 2/174) and K. pneumoniae (2.0%, 2/99) isolates separately. For the Enterobacterales, CA was mostly lowered by a 6.7% (52/775) minor error rate.
FIG 2.
Frequency tables for Enterobacterales (A), P. aeruginosa (B), and A. baumannii (C) categorized according to CLSI-FDA and EUCAST breakpoints. Blue lines, EUCAST breakpoint; red lines, CLSI breakpoint; green lines, EUCAST and CLSI breakpoints. a, CLSI/EUCAST interpretive criteria; b, two isolates excluded from the evaluation according to ISO criteria because the variability was too high; c, changes in BMD values due to ISO resolution of MEs and VMEs.
Among the P. aeruginosa isolates, EA and CA were 98.3% (117/119) and 93.3% (111/119), respectively, meeting the criteria for both (Fig. 2B). Using A. baumannii complex isolates, the P/T Etest met the criteria for EA (91.6%, 76/83) but not those for CA (89.2%, 74/83) (Fig. 2C). Neither VMEs nor MEs were found among the P. aeruginosa and A. baumannii complex isolates. Categorical agreement was lowered for the P. aeruginosa and A. baumannii complex isolates by minor error rates that were 6.7% (8/119) and 10.8% (9/83), respectively.
Performance compared to ISO requirements.
ISO 20776-2 standard performance criteria (≥90.0% EA and/or CA, ≤3.0% VMEs or MEs, ≥95.0% reproducibility, results for ≥95.0% of the QC isolates within the expected range) were also used to evaluate the P/T Etest using EUCAST breakpoints (Table 3). One E. coli isolate and one K. pneumoniae isolate were excluded from the EUCAST-ISO evaluation because of high variability in the BMD results per the directions for the ISO 20776-2 standard. For the Enterobacterales, EA and CA were 96.6% (747/773) and 94.8% (733/773), respectively (Fig. 2A). Among the Enterobacterales, no VMEs (0/224) were found, but 2 MEs (0.4%, 2/525) were observed: one E. coli isolate had MIC values of 32 μg/ml and 4 μg/ml by the P/T Etest and BMD, respectively, and one Morganella morganii isolate had an MIC result of 32 μg/ml by the P/T Etest and an MIC result of 1 μg/ml by BMD. For the Enterobacterales, CA was mostly lowered by a 4.9% (38/773) minor error rate.
For P. aeruginosa, EA and CA were 98.3% (117/119) and 95.8% (114/119), respectively, meeting the ISO criteria. However, for this species, VME and ME rates were, respectively, 6.9% (2/29) and 3.3% (3/90), both of which are higher than the ISO criteria (Fig. 2B). These 5 categorical errors originated from isolates with Etest results 1 doubling dilution apart from the BMD results, therefore falling in the criterion for EA.
Reproducibility study.
The reproducibility study included 4 P. aeruginosa isolates, 3 E. coli isolates, 2 K. pneumoniae isolates, 1 M. morganii isolate, and 2 A. baumannii isolates. Evaluation in triplicate for 3 days resulted in 108 determinations per site and 324 in total. All determinations fell within ±1 doubling dilution of the mode, so reproducibility met the criteria for best- and worst-case calculations and was 100% for both between-site (270/270) and within-site (90/90) reproducibility, according to both FDA and ISO requirements (Table 4).
TABLE 4.
Reproducibility of the new P/T Etest
| Organism | No. of isolates with the following no. of dilutions different from test mode |
MIC (mg/liter) by test mode | ||||||
|---|---|---|---|---|---|---|---|---|
| Off scale | −2 | −1 | 0 | 1 | 2 | Off scale | ||
| A. baumannii | 12 | 15 | 32 | |||||
| A. baumannii | 27 | 16 | ||||||
| E. coli | 15 | 12 | 32 | |||||
| E. coli | 7 | 20 | 32 | |||||
| E. coli | 5 | 22 | 16 | |||||
| K. pneumoniae | 27 | 2 | ||||||
| K. pneumoniae | 9 | 17 | 1 | 2 | ||||
| M. morganii | 13 | 14 | 16 | |||||
| P. aeruginosa | 6 | 18 | 3 | 128 | ||||
| P. aeruginosa | 19 | 8 | 16 | |||||
| P. aeruginosa | 23 | 4 | 16 | |||||
| P. aeruginosa | 22 | 5 | 2 | |||||
| Total | 0 | 0 | 52 | 239 | 33 | 0 | 0 | |
Quality control.
A total of 85 QC replicates were performed among the 4 sites. One hundred percent of the results were within the expected ranges defined by the CLSI M100 (13) and EUCAST (14) standards for both the Etest and BMD, meeting both FDA and ISO requirements.
DISCUSSION
P/T is a widely used β-lactam–β-lactamase inhibitor combination for treating several complicated infections, as recognized in different treatment guidelines. Recently, a renewed interest in the use of this combination against ESBL producers has emerged, particularly to avoid the use of carbapenems in order to limit the spread of carbapenemase-producing bacteria (4–6, 17).
On the other hand, routine clinical laboratories also use other methods as alternatives to the standard BMD to obtain MIC values. These methods need to be easy to perform and must be properly evaluated according to performance standards (18). One of these methods is diffusion with gradient strips, which also included P/T strips. However, bioMérieux, one of the manufacturers of the gradient test, decided to discontinue production of the Etest strip for P/T (formerly called Etest PTc) concomitantly with a communication from the EUCAST Development Laboratory alerting about the inaccuracy of the susceptibility testing results obtained with these strips (http://www.eucast.org/ast_of_bacteria/warnings/). Moreover, in the MERINO trial, some of the MIC values were obtained with the withdrawn Etest PTc strip, and the results were used to establish correlations with clinical outcomes (17). This fact created controversies due to potential interpretive errors (a high VME rate), and further analyses of the isolates involved in this trial using BMD have been performed (19).
After the EUCAST warning, the piperacillin-tazobactam strip was redeveloped and evaluated in the large multicenter trial summarized in this paper. Results from this evaluation were used for the CE (December 2018) and FDA (January 2019) approvals. To the best of our knowledge, this is the first study comparing the performance for this new P/T Etest strip to that of the standard BMD. Similar evaluations have recently been performed for ceftolozane-tazobactam and ceftazidime-avibactam MIC strips (20, 21).
In our study, we obtained a good overall performance with this new P/T Etest, with EA values being higher than 90% for Enterobacterales, P. aeruginosa, and A. baumannii complex isolates, exhibiting an excellent correlation between Etest and BMD according to both EUCAST-ISO and CLSI-FDA recommendations. Two VMEs corresponding to two clinical K. pneumoniae isolates were observed in the evaluation performed according to FDA recommendations. Only one of those two VMEs was reproduced when the isolates were retested. Indeed, for the second isolate, BMD highlighted a variability too large to conclude its resistance to piperacillin-tazobactam with confidence.
In the evaluation performed according to EUCAST-ISO recommendations, 3 (3.3%) MEs and 2 (6.9%) VMEs were observed among the P. aeruginosa isolates, resulting in the lack of achievement of the ME/VME rates for this microorganism. However, these results need to be analyzed by consideration of the lack of an intermediate category for P/T with P. aeruginosa. As a consequence, when the reference result is just on the breakpoint MIC value, 1 dilution apart from the Etest result, either a ME or a VME is produced. Indeed, the 5 categorical errors originated from isolates with Etest-based MIC results 1 doubling dilution apart from the BMD results, which were, consequently, in the EA category.
The strengths of this study include the use of a collection of isolates with different mechanisms of resistance to different β-lactam antibiotics (see Table S1 in the supplemental material), the large number of clinical isolates obtained, a multicenter evaluation comprising three tertiary care medical centers in different geographic areas, and the use of BMD as a reference method with a standardized and validated preparation of panels. The limitations include the lack or a limited number of resistant isolates of specific species. Indeed, no resistant Proteus mirabilis, Proteus vulgaris, or Salmonella enterica isolates were found when performing the trial. Despite these limitations, the results of this trial resulted in both the CE marking and the FDA clearance of the P/T Etest. In addition, using the P/T Etest, the EUCAST Development Laboratory generated data leading to the lifting of the warning on the piperacillin-tazobactam Etest that had been issued in 2015.
In conclusion, compared to the results of the reference BMD method, the results of this multicenter trial support the accuracy of the new P/T Etest for determining the MICs for Enterobacterales, P. aeruginosa, and A. baumannii complex isolates in the clinical setting.
Supplementary Material
ACKNOWLEDGMENTS
We acknowledge the great work of Johanne Gafsi, Marissa Loubet, and Karen Macdonald from bioMérieux during the undertaking of this trial.
This study was funded by a research grant from bioMérieux (Marcy l’Etoile, France) and was partially supported by the Plan Estatal de I+D+i 2013–2016 and the Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía, Industria y Competitividad, Spanish Network for Research in Infectious Diseases (grant RD16/0016/0011), cofinanced by the European Development Regional Fund (ERDF; A way to achieve Europe), Operative Program Intelligent Growth 2014–2020. S.G.-F. is supported by a research contract from the Instituto de Salud Carlos III, Spain (Rio Hortega program, grant CM17/00033). Funding for the study performed at the Washington University study site was provided by bioMérieux.
R.C. has participated in educational programs organized by bioMérieux. Y.B., T.A., and G.Z. are employees of bioMérieux. The other authors have no conflict of interest.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Schoonover LL, Occhipinti DJ, Rodvold KA, Danziger LH. 1995. Piperacillin/tazobactam: a new beta-lactam/beta-lactamase inhibitor combination. Ann Pharmacother 29:501–514. doi: 10.1177/106002809502900510. [DOI] [PubMed] [Google Scholar]
- 2.Drawz SM, Bonomo RA. 2010. Three decades of beta-lactamase inhibitors. Clin Microbiol Rev 23:160–201. doi: 10.1128/CMR.00037-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bush K, Bradford PA. 2019. Interplay between β-lactamases, and new β-lactamase inhibitors. Nat Rev Microbiol 17:295–306. doi: 10.1038/s41579-019-0159-8. [DOI] [PubMed] [Google Scholar]
- 4.Harris PNA, Tambyah PA, Paterson DL. 2015. β-Lactam and β-lactamase inhibitor combinations in the treatment of extended-spectrum β-lactamase producing Enterobacteriaceae: time for a reappraisal in the era of few antibiotic options? Lancet Infect Dis 15:475–485. doi: 10.1016/S1473-3099(14)70950-8. [DOI] [PubMed] [Google Scholar]
- 5.Tamma PD, Rodriguez-Bano J. 2017. The Use of noncarbapenem β-lactams for the treatment of extended-spectrum β-lactamase infections. Clin Infect Dis 64:972–980. doi: 10.1093/cid/cix034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rodríguez-Baño J, Navarro MD, Retamar P, Picón E, Pascual Á, Extended-Spectrum Beta-Lactamases–Red Española de Investigación en Patología Infecciosa/Grupo de Estudio de Infección Hospitalaria Group. 2012. β-Lactam/β-lactam inhibitor combinations for the treatment of bacteremia due to extended-spectrum β-lactamase-producing Escherichia coli: a post hoc analysis of prospective cohorts. Clin Infect Dis 54:167–174. doi: 10.1093/cid/cir790. [DOI] [PubMed] [Google Scholar]
- 7.Tamma PD, Han JH, Rock C, Harris AD, Lautenbach E, Hsu AJ, Avdic E, Cosgrove SE, Antibacterial Resistance Leadership Group. 2015. Carbapenem therapy is associated with improved survival compared with piperacillin-tazobactam for patients with extended-spectrum β-lactamase bacteremia. Clin Infect Dis 60:1319–1325. doi: 10.1093/cid/civ003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gutiérrez-Gutiérrez B, Rodríguez-Baño J. 2019. Current options for the treatment of infections due to extended-spectrum beta-lactamase-producing Enterobacteriaceae in different groups of patients. Clin Microbiol Infect 25:932–942. doi: 10.1016/j.cmi.2019.03.030. [DOI] [PubMed] [Google Scholar]
- 9.ISO. 2006. ISO/FDIS 20776-1:2006(E) Clinical laboratory testing and in vitro diagnostic test systems. Susceptibility testing of infection agents and evaluation of performance of antimicrobial susceptibility test devices—part 1, reference method for testing the in vitro activity of antimicrobial agents against rapidly growing aerobic bacteria involved in infectious diseases. ISO, Geneva, Switzerland. [Google Scholar]
- 10.Brown DF, Brown L. 1991. Evaluation of the E test, a novel method of quantifying antimicrobial activity. J Antimicrob Chemother 27:185–190. doi: 10.1093/jac/27.2.185. [DOI] [PubMed] [Google Scholar]
- 11.Clinical and Laboratory Standards Institute. 2018. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 11th ed CLSI standard M07-A10 Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 12.FDA. 2009. FDA guidance for industry: class II special controls guidance document: antimicrobial susceptibility test (AST) systems. FDA, Silver Spring, MD. [Google Scholar]
- 13.Clinical and Laboratory Standards Institute. 2018. Performance standards for antimicrobial susceptibility testing. 28th ed CLSI supplement M100 Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 14.European Committee on Antimicrobial Susceptibility Testing. 2018. Routine and extended internal quality control for MIC determination and disk diffusion as recommended by EUCAST, version 8.0. http://www.eucast.org.
- 15.European Committee on Antimicrobial Susceptibility Testing. 2018. Breakpoint tables for interpretation of MICs and zone diameters, version 8.0. http://www.eucast.org.
- 16.ISO. 2007. ISO/FDIS 20776–2:2007(E) clinical laboratory testing and in vitro diagnostic test systems. Susceptibility testing of infection agents and evaluation of performance of antimicrobial susceptibility test devices—part 2, evaluation of performance of antimicrobial susceptibility test devices. ISO, Geneva, Switzerland. [Google Scholar]
- 17.Harris PNA, Tambyah PA, Lye DC, Mo Y, Lee TH, Yilmaz M, Alenazi TH, Arabi Y, Falcone M, Bassetti M, Righi E, Rogers BA, Kanj S, Bhally H, Iredell J, Mendelson M, Boyles TH, Looke D, Miyakis S, Walls G, Al Khamis M, Zikri A, Crowe A, Ingram P, Daneman N, Griffin P, Athan E, Lorenc P, Baker P, Roberts L, Beatson SA, Peleg AY, Harris-Brown T, Paterson DL, for the MERINO Trial Investigators and the Australasian Society for Infectious Disease Clinical Research Network (ASID-CRN). 2018. Effect of piperacillin-tazobactam vs meropenem on 30-day mortality for patients with E coli or Klebsiella pneumoniae bloodstream infection and ceftriaxone resistance: a randomized clinical trial. JAMA 320:984–994. doi: 10.1001/jama.2018.12163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.EU. 2017. Regulation (EU) 2017/746 of the European Parliament and of the Council of 5 April 2017 on in vitro diagnostic medical devices and repealing directive 98/79/EC and Commission Decision 2010/227/EU. Offic J Eur Union 117:176e331. [Google Scholar]
- 19.Henderson A, Tambyah PA, David C, Lye DC. 2019. Association with 30-day mortality and MIC in patients treated with piperacillin/tazobactam for Escherichia coli and Klebsiella pneumoniae bloodstream infections that are non-susceptible to ceftriaxone from patients enrolled in the MERINO trial, abstr P-2468 Abstr 29th Eur Congr Clin Microbiol Infect Dis, Amsterdam, Netherlands European Society of Clinical Microbiology and Infectious Diseases, Basel, Switzerland. [Google Scholar]
- 20.Bailey AL, Armstrong T, Dwivedi H-P, Denys GA, Hindler J, Campeau S, Traczewski M, Humphries R, Burnham C-A. 2018. Multicenter evaluation of the Etest gradient diffusion method for ceftolozane-tazobactam susceptibility testing of Enterobacteriaceae and Pseudomonas aeruginosa. J Clin Microbiol 56:e00717-18. doi: 10.1128/JCM.00717-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kresken M, Körber-Irrgang B. 2018. Performance of the Etest for susceptibility testing of Enterobacterales (Enterobacteriaceae) and Pseudomonas aeruginosa toward ceftazidime-avibactam. J Clin Microbiol 56:e00528-18. doi: 10.1128/JCM.00528-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.